Agarose beads are the most common base matrix for chromatography media. Agarose is ideal as a base matrix because of its minimal non-specific absorption, hydrophilicity, strong chemical resistance to acid, base and solvents, high porosity and abundance of OH groups for crosslinking and functionalization. Most ion-exchange and affinity packed-bed media are based on agarose beads. Another type of media is particles with an agarose coating on the outside. Such particles are useful for fluidizing bed because the density of those particles must be controlled to counter the buoyancy of the fluidizing flow.
The most common method of making agarose beads is by contacting an aqueous liquid and a hydrophobic liquid in a stirring vessel. This batch process can be used for making both homogeneous and cored beads. In the case of homogeneous beads, agarose solids are dissolved in water heated to about 90° C. The hot solution is then poured into a hot hydrophobic fluid in a stirring vessel. The hydrophobic fluid can be a solvent such as toluene or mineral oil. Since the two fluids don't mix, constant agitation turns the two liquids into an emulsion with the agarose solution as droplets suspended in the hydrophobic fluid. Normally, a surfactant soluble in the phobic fluid is added to stabilize the droplets so they don't coalesce into larger ones. The emulsion is then cooled to cause the agarose beads to gel. The solids are then washed and sieved to narrow the distribution to the useful range.
U.S. Pat. No. 4,971,833 and U.S. Pat. No. 5,866,006 use this same process to produce cored beads. The only difference is mixing cores with the agarose solution before mixing with the phobic fluid. Agitation breaks up the agglomeration of cores and agarose solution into smaller units. After a certain residence time of about 5-10 minutes, the solution is cooled down gradually generally in about 30 minutes to solidify the agarose solution into a gel. The solids are then washed and sieved to narrow the distribution to the useful range.
For homogeneous beads, the limitation of the batch process is the throughput. In order for the agarose to form droplets, the volume of the phobic phase has to be at least 3 to 1 that of the agarose solution. For example, to prepare 500 L of beads, a vessel must contain a total of 2000 L of solvent and agarose solution. In large operations, it is not uncommon to use reactors with several thousand of gallons capacity. Big vessels like this take much longer time to heat and cool. In addition even with the most optimal settings, the distribution of the bead size is wide and low single core would result. Further, mixing inside a stirring vessel is inherently a statistical process. Each bead follows its own path and each one has a different shear history. Due to these non-uniformities, the resulting beads can have widely different properties in terms of size and/or coating thickness, etc.
Most patents describe bench scale processes in which the whole process, heating and cooling, happen within 30 minutes. In actual mass production scale, the process time could be much longer. Additionally, since a flammable solvent such as toluene is used most of the time, explosion-proofing the equipment and facilities can get very expensive.
There are additional problems using this process to produce cored beads. The shear forces are non-uniform in a mixing vessel. In the dead areas of the vessel, shear forces are not sufficient to break up the cores, thus resulting in large agglomerations of multiple cores. However, if the stirring speed is too high, the shear forces near the impeller may strip the agarose solution partially and/or completely away from the core. For this reason, a mixing vessel usually produces a low yield of single core beads. A majority of the beads have two or more cores. While multiple cores may be desirable for fluidizing bed to increase their density and control their movement in the bed, See U.S. Pat. No. 6,428,707, it is not the preferred population for packed bed chromatography in which control of the diffusion path and rigidity are the ultimate goals.
Instead of a bulky batch process, it is desirable to use a truly continuous process for bead manufacturing. Although batch agarose processing are widely researched and documented, patents and literatures are silent when it comes to continuous processing.
Therefore, it is highly desirable to develop a process that has high throughput, high single core yield, is continuous, does not require explosion proof equipment, is compact in size and inexpensive to build and operate. To provide further flexibility, it is also desirable to use the same process to produce either homogenous or cored beads. The present invention provides such a process.